Prediction Markets, AI Vulnerabilities, and Quantum Threats to Crypto

Source: Blockworks Original Title: Thursday links: Prediction markets, agent hackers, quantum risks Original Link: https://blockworks.co/news/agent-hackers-quantum-risks

The Evolution of Prediction Markets

Michael Lewis traces the origin of proposition betting to 1985 when Caesars Palace offered 20-1 odds on William “Refrigerator” Perry scoring a touchdown in the Super Bowl. Perry did score, causing Caesars to lose at least $250,000—yet the casino president called it their best bet ever due to the publicity generated.

This sparked a revolution in betting. Before Perry’s touchdown, there were only about three bets available on the Super Bowl: the winner, total points, and halftime score. Today, there are hundreds of proposition bets, and bookmakers now write more business on props than on the games themselves.

This evolution laid the foundation for today’s sophisticated prediction markets, which now offer odds on virtually anything.

Political Markets and Their Feedback Loops

Prediction markets have become so sophisticated that some traders check them seconds before major events occur—markets often move ahead of traditional media feeds.

Stanford political science professor Andy Hall sees prediction markets as potentially offering “a clearer shared picture of a highly complex political environment.” However, he warns of concerning feedback loops. During a Virginia Attorney General race, unverified exit poll claims on social media moved prediction markets, which then became “breaking news,” further amplifying market movements.

Hall also raises questions about what it means to “win” an election when prediction markets are treated as sources of truth. If prediction markets “call” a close presidential election in 2028, the implications could be profound.

AI Agents as Smart Contract Hackers

Anthropologic researchers report that AI agents successfully exploited 56% of smart contracts with known vulnerabilities. More alarming, when tested on 2,849 smart contracts with no known vulnerabilities, the agents discovered two novel zero-day exploits.

The rate of improvement is staggering: frontier models’ exploit revenue doubled roughly every 1.3 months over the past year—compared to Moore’s Law’s semiconductor doubling every two years.

Cost is another concern: scanning a contract for vulnerabilities costs just $1.22 on average. As costs continue falling, attackers will deploy more AI agents to probe any code along the path to valuable assets.

Quantum Computing and Crypto Vulnerability

While quantum threats to cryptography are often dismissed as a future problem affecting everything equally, crypto faces unique vulnerabilities.

Classical cryptography researchers have been working on quantum resistance for years and already have protocols ready. Web2 systems have centralized authorities that could halt transactions if needed. Crypto, however, faces two critical problems:

1. Blockchains use novel cryptography applications that have received little attention from quantum-resistance researchers
2. The decentralized nature of crypto makes infrastructure changes extremely challenging

Moreover, blockchains were specifically designed to make data available to everyone at all times—meaning quantum computers could decrypt entire transaction histories of every blockchain.

Privacy coins like Monero are already considered vulnerable. Governments could use quantum computers to identify unreported crypto gains and send tax bills accordingly. Satoshi’s coins could be moved, and rival nations could target Bitcoin not for profit but to destabilize a system increasingly integrated into US finance.

Despite a $3 trillion crypto market capitalization, investment in quantum-resistant solutions remains minimal. Experts estimate that a 1% risk of total loss should warrant $30 billion in defensive investment—yet current spending is essentially zero.